A medical experiment where neither the patient or the doctor knows who has the active drug and who has the placebo.
This design is deemed superior to a single blind trial, as the results can in no way be affected by the knowledge of the placebo.